Issue metadata
Sign in to add a comment
|
Scroll jumped to top when playing transition animation that includes transform: scale().
Reported by
johnson....@appier.com,
Oct 11 2016
|
||||||||||||||||||||||||
Issue descriptionUserAgent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_11_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/53.0.2785.143 Safari/537.36 Steps to reproduce the problem: 1. Visit http://jsbin.com/vuseqav/4/edit?html,css,output 2. In the "Output" panel, scroll down a bit until "mouse-over me" is in the top of the screen 3. Hover on the "mouse-over me" What is the expected behavior? A tomato-colored square with "Animated item" appears after a scaling transition animation. http://i.giphy.com/l2Sq1yIjVQDso4il2.gif What went wrong? The surrounding text jumped down (with a distance that seems related to how much we have scrolled down) a bit when the scaling animation is playing, and jumped back when after the animation ends. http://i.giphy.com/26ufktHmzfbQfQkj6.gif Did this work before? Yes Chrome 52 Does this work in other browsers? Yes Chrome version: 53.0.2785.143 Channel: stable OS Version: OS X 10.11.6 Flash Version: Shockwave Flash 23.0 r0 * If `will-change: transform` is added ( http://jsbin.com/kokabit/1/edit?html,css,output ), the scrolling behavior becomes super wierd ( http://i.giphy.com/26uf6gRKsjbNy463e.gif ) -- texts are disappearing when scrolling. I believe it is a bug in GPU acceleration. * There is an issue in material-design-lite that is probably related to this: https://github.com/google/material-design-lite/issues/4574#issuecomment-246600670 (starting from the 2nd comment) * In Google Chrome Canary 56.0.2886.0 canary (64-bit) the problem does not exist anymore. However, I think there is a serious issue in the render engine that we cannot wait for a few months to be resolved.
,
Oct 11 2016
,
Oct 12 2016
As per comment #1 tested this issue in chrome beta version #54.0.2840.0 below are the good and bad builds in Win10.0 ,Mac 10.12 & Ubuntu 14.04. Unable to provide tool bisect using below versions, so providing manual change log below Good Build: 54.0.2840.33 (not observed the blink) Bad Build: 54.0.2840.32 (observed blink) Manual CL as per omahaproxy: -------------------------------- https://chromium.googlesource.com/chromium/src/+log/54.0.2840.32..54.0.2840.33?pretty=fuller&n=10000 Unable to find possible suspect from the above CL. Could anyone please help us to find the right owner Note: Not able to reproduce on Canary #56.0.2887.0/ Dev #55.0.2883.9/ Beta #54.0.2840.50 Thank you.
,
Oct 13 2016
Manually bisected non-release builds. Unregression range: https://chromium.googlesource.com/chromium/src/+log/9aba30b71b698e84fbbbd10972683b6c7c627fb0..fb7bfc94a65058ccf120e1db58ceae7d7c6bf526 Suspecting https://chromium.googlesource.com/chromium/src/+/578237496c6f4615f752e1c30e50971d118bf208 fixed the bug. chrishtr: Please consider merging this change to stable if it fixes this bug.
,
Oct 14 2016
I verified that the commit referenced in comment 4 indeed fixed the bug. Marking as a duplicate of 646363. I tried to get approval to push the fix for 646373 to M53 previously, but it was too rocky of a release. At this point we should just wait for M54. |
|||||||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||||||
Comment 1 by alancutter@chromium.org
, Oct 11 2016Owner: alancutter@chromium.org
Status: Available (was: Unconfirmed)